Mitochondrial form and function in pancreatic β-cells and brown adipocytes - Jakob D Wikström, M.D - DIVA
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
© Jakob D Wikström, Stockholm 2010 Picture on cover: Brown adipocyte stained with the red mitochondrial membrane potential sensitive dye TMRE. In addition a subset of the mitochondrial population is also labeled with photo-convertible GFP. The image was acquired with a Zeiss LSM 710 confocal microscope. ISBN 978-91-7447-095-6 Printed in Sweden by Universitetsservice AB, Stockholm 2010 Distributor: Stockholm University Library
“If I could remember the names of all these particles I’d be a botanist” Albert Einstein “Nothing shocks me. I’m a scientist.” Indiana Jones “Art is I; Science is we” Claude Bernard To my parents
Abstract Mitochondria stand in the center of metabolism and compromised mitochondrial function has been shown in metabolic diseases such as diabetes and obesity. To develop mitochondrial targeted therapeutics, an improved understanding of the regulation of mitochondrial function is needed. This thesis is focused on the role of mitochondria in two tissues highly dependent on mitochondria: pancreatic β-cells and brown adipose tissue (BAT). The role of mitochondria in these tissues is opposite. In β-cells, mitochondrial ATP production is necessary for insulin secretion while in BAT mitochondria produce heat by disconnecting the respiratory chain from ATP synthesis. Two main aspects of mitochondria were explored; mitochondrial functional efficiency and the interrelationship between mitochondrial shape and function. Mitochondria within dispersed individual β-cells were found to exhibit heterogeneity in mitochondrial membrane potential. This functional diversity decreased when cells were challenged with glucose stimuli, suggesting that at higher fuel levels low-activity mitochondria are recruited into a pool of high-activity mitochondria. The BCL-2 family member BAD was identified as a contributor to mitochondrial membrane potential heterogeneity. Glucolipotoxic conditions designed to mimic diabetes in vitro increased the functional diversity suggesting that this may be of importance for diabetes pathophysiology. To examine mitochondrial efficiency in intact islets a high throughput islet respirometry method was developed. It was found that due to increased uncoupling, islets from a diabetic animal model exhibit lower respiratory efficiency as compared to animals fed control chow. Glucose, free fatty acids and amino acids all decreased respiratory efficiency. A large portion of the respiratory efficiency appeared mediated by reactive oxygen species and the adenine nucleotide translocase. Human islets showed higher respiratory efficiency as compared to mouse islets. However, as in the mouse islets glucose decreased respiratory efficiency. In islets from obese donors there was a trend towards decreased respiratory rates. The interrelationship between mitochondrial shape and function was examined both in β-cells and BAT. In β-cells mitochondria were found to undergo cycles of fusion and fission. The key mitochondrial dynamics proteins Opa1, Drp1, and Fis1 were shown to regulate β-cell mitochondrial morphology. During glucolipotoxicity mitochondria fragmented and lost their fusion ability. Knock down of the fission protein Fis1 rescued the β-cells from glucolipotoxic induced cell death and maintained β-cell insulin secretion capacity. Similarly, BAT mitochondria also showed fusion and fission. The mitochondrial dynamics proteins Mfn2 and Drp1 were shown to strongly affect BAT mitochondrial morphology. In response to a combination of adrenergic and free fatty acid stimuli mitochondria drastically changed from long filamentous structures to fragmented spheres. This occurred as a wave passing through the cell. The mitochondrial fragmentation was dependent on the β-adrenergic pathway and reactive oxygen species. Inhibiting fission by the negative form of Drp1 decreased BAT response to adrenergic stimuli by half. Thus, mitochondrial fission appeared essential for proper BAT function. In conclusion, mitochondrial efficiency may be of importance for normal as well as compromised β-cell and islet function. Mitochondrial morphology appears critical for mitochondrial function in β-cells and BAT.
This thesis is based on the following papers, which are referred to in the text by their Roman numerals, respectively. I. Wikstrom JD, Katzman SM, Mohamed H, Twig G, Graf SG, Heart E, Corkey BE, de Vargas LM, Danial NN, Collins S, Shirihai OS. Mitochondrial functional heterogeneity in pancreatic beta cells. Diabetes: 56(10):2569-78, 2007. II. Molina AJ, Wikstrom JD, Stiles L, Las G, Mohamed H, Elorza A, Walzer G, Twig G, Katz S, Corkey BE, Shirihai OS. Mitochondrial Networking Protects Beta Cells from Nutrient Induced Apoptosis. Diabetes. 58(10):2303-15, 2009. III. Wikstrom JD, Elorza A, Allister EM, Sereda SB, Stiles L, Neilson A, Neilson A, Ferrick DA, Corkey BE, Deng S, Wheeler MB, Shirihai OS. Islet bioenergetic efficiency is regulated by nutrients. Manuscript IV. Wikstrom JD, Si Y, Twig G, Liesa M, Sereda SB, Las G, Cannon B, Nedergaard J, Shirihai OS. Brown adipocyte activation is characterized by a wave of mitochondrial fission and depolarization that is dependent on β3 receptor stimulation and Drp1, and is characterized by complete, but reversible, arrest of fusion Manuscript It also refers to the following papers. Wikstrom JD, Twig G, Shirihai OS. What can mitochondrial heterogeneity tell us about mitochondrial dynamics and autophagy? Intern J. Biochem Cell Biol. 41(10):1914-27, 2009. Danial NN, Walensky LD, Zhang CY, Choi CS, Fisher JK, Molina AJ, Datta SR, Pitter KL, Bird GH, Wikstrom JD, Deeney JT, Robertson K, Morash J, Kulkarni A, Neschen S, Kim S, Greenberg ME, Corkey BE, Shirihai OS, Shulman GI, Lowell BB, Korsmeyer SJ. Dual role of pro-apoptotic BAD in insulin secretion and beta cell survival. Nat. Med. 14(2):144-53, 2008. Katzman SM, Messerli MA, Barry DT, Grossman A, Harel T, Wikstrom JD, Corkey BE, Smith PJ, Shirihai OS. Mitochondrial metabolism reveals a functional architecture in intact islets of Langerhans from normal and diabetic Psammomys obesus. Am J Physiol Endocrinol Metab 287(6):1090-1099, 2004. Twig G, Graf SA, Wikstrom JD, Mohamed H, Haigh SE, Elorza A, Deutsch M, Zurgil N, Reynolds N, and Shirihai OS. Tagging and tracking individual networks within a complex mitochondrial web with photoactivatable GFP. Am J Physiol Cell Physiol 291, C176-C184, 2006.
Twig G; Liu X; Liesa M; Wikstrom JD; Molina AJA; Las G; Yaniv G; Hajnoczky G; Shirihai OS. Biophysical properties of mitochondrial fusion events in pancreatic β-cells and cardiomyocytes unravel potential control mechanisms of its selectivity. Am J Physiol Cell Physiol, May 8 2010. [Epub ahead of print]
Contents 1. Introduction 12 2. Mitochondrial physiology 13 2.1. Mitochondria in a nut shell 13 2.2. Principles of bioenergetics 14 2.3. Mitochondrial oxygen consumption in cells and pancreatic islets 15 2.3.1. Assessing mitochondrial function 15 2.3.2. Uncoupled and coupled respiration 15 2.3.3. Islet respiration 15 2.3.4. Significance of measuring islet respiration 16 2.3.5. Uncoupled respiration of islets 17 2.3.6. Significance of uncoupled respiration 17 2.3.7. Regulation of uncoupled respiration 18 2.3.7.1. Fuel regulation of uncoupled respiration 18 2.3.7.2. Lipid composition 19 2.3.7.3. Uncoupling proteins 19 2.3.7.4. Adenine nucleotide translocase 20 2.3.7.5. Permeability transition pore 20 2.3.7.6. Reactive oxygen species 21 2.4. Mitochondrial dynamics . 22 2.4.1 Regulation of mitochondrial fusion 22 2.4.2. Regulation of mitochondrial fission 22 3. Mitochondrial heterogeneity 23 3.1 Subcellular heterogeneity 24 3.1.1. Mitochondrial membrane potential heterogeneity 24 3.1.2. JC-1 25 3.1.3. TMRE/TMRM allows for quantitative evaluation. 26 3.2. Heterogeneity in compromised cells 26 3.3 Mechanism of mitochondrial heterogeneity 27 3.3.1 Subcellular location 27 3.3.1.2. Perinuclear vs. peripheral locations 27 3.3.1.3. Calcium 28 3.3.1.4. Access to metabolites 28 3.3.1.5. F1F0-ATPase 29 3.3.2. Diversity vs. instability 29 3.3.3. Organelle content 30 3.4. How does mitochondrial heterogeneity coexist with mitochondrial 31 dynamics? 3.4.1. Fusion—not for everyone 32 3.4.2. Mitochondrial “kiss and run” generates heterogeneity 32 3.4.3. Do fusion events result in complete equilibration? 33 3.4.3.1. Matrix 33 3.4.3.2. Membranes 33
3.4.3.3. mtDNA 34 3.4.3.4. Inner membrane vs. matrix mixing 35 3.5. Autophagy, a mechanism that removes depolarized mitochondria 35 3.5.1. Mitophagy reduces mitochondrial heterogeneity 36 3.5.2. “Death row” mitochondria contribute to heterogeneity 36 3.5.3. Tags for selective mitophagy requires heterogeneity 37 3.5.3.1. Characteristics of mitochondria in the pre-autophagic pool 37 4. Mitochondrial pathophysiology in β-cells and islets 39 4.1. Insulin Resistance 39 4.2. β-cell mitochondrial dysfunction 39 4.3. Glucolipotoxicity 39 4.4. Mitochondrial morphology and dynamics in diabetes 40 4.5. Rodent islet mitochondrial function 41 4.6. Human islet mitochondrial function 42 5. Mitochondria in brown adipose tissue 43 5.1. BAT function and significance 43 5.2. UCP1 and BAT signaling 44 5.3. ROS in BAT 45 5.4. UCP1 content 46 5.5. Mitochondrial membrane potential 46 5.6. Mitochondrial morphology in BA 48 5.7. Mitochondrial dynamics in brown adipocytes 50 6. Conclusions and summary of thesis 51 6.1. Thesis summary 51 6.2. Conclusions and future perspectives 52 7. Acknowledgements 54 8. References 56
Abbreviations Body mass index (BMI) Brown adipocytes (BA) Brown adipose tissue (BAT) Dynamin-related protein 1 (Drp1) Free fatty acids (FFA) Green fluorescent protein (GFP) Inner membrane PAGFP(IMPAGFP) Inner mitochondrial membrane (IM) Insulin growth factor 1 receptor (IGF1R) Matrix Red flurescent protein (MTRFP) Matrix targeted PA-GFP (MTPAGFP) Mitochondrial DNA (mtDNA) Mitochondrial membrane potential (ΔΨm) Mitofusin (Mfn) Outer membrane GFP (OMGFP) Photo-activatable GFP (PA-GFP) Polyethylene glycol (PEG) Rat insulinoma derived beta cell line (INS1) Reactive oxygen species (ROS) Short interference RNA interference (siRNAi) Tetrachloro-1,1,3,3-tetraethylbenzimidazol-carbocyanineiodide (JC-1) Tetrakis (4-benzoic acid) porphyrin manganese(III) (TBAP) Type 2 diabetes (T2D) Uncoupling protein 2 (UCP2)
1. Introduction Type 2 diabetes and obesity are metabolic diseases that plague the world in the 21st century. In the United States alone, it is now estimated that 10% of adults fcare diabetic (Centers for Disease Control and Prevention, 2007) and 35% are obese (Catenacci, 2009). In Sweden, 292,000 people were diabetic in 2000 (World health organization, 2010). According to the World Health Organization it affected 171 million people in 2001 and by the year 2030 is expected to grow to a staggering 366 million. Diabetes is the leading cause of blindness, end-stage renal disease and neuropathy in the US (Brownlee, 2003). The direct and indirect costs related to diabetes in the U.S. are estimated at $132 billion in medical expenditures and lost productivity (Hogan, 2003). Although significant research effort has been invested, there has been a poor delivery in terms of therapeutics. Thus, more research is needed. Normoglycemia is maintained by several tissues including pancreatic islets and skeletal muscle. Diabetes is defined as elevated blood glucose; hyperglycemia. In type 2 diabetes tissues are considered to malfunction. Skeletal muscle display insulin resistance, i.e. low uptake of blood glucose. Pancreatic β-cell insulin secretion is insufficient or dysregulated. Of great importance for normal function of these tissues are mitochondria (Maechler, 2006). There is a growing body of data indicating that β-cell mitochondria malfunction in type 2 diabetes (Mulder, 2009). This is perhaps not surprising considering that mitochondria have a central role in metabolism. The thermogenic brown adipose tissue (BAT) is possibly the organ that is most characterized by its mitochondria. This because the heat produced is a direct product of mitochondrial activity (Cannon, 2004). It has been suggested that BAT may play an important role in metabolic disease as it transforms stored energy into heat when active. A recent study indicates that BAT may be dysfunctional in this category of patients (Cypess, 2009). Mitochondria take up a substantial portion of the cytoplasmic volume of eucaryotic cells, and they have been essential for the evolution of complex animals. Without mitochondria, present-day animal cells would be dependent on anaerobic glycolysis for all of their ATP. When glucose is converted to pyruvate by glycolysis, only a very small fraction of the total free energy potentially available from the glucose is released. In mitochondria, the metabolism of sugars is completed: the pyruvate is imported into the mitochondrion and oxidized by O2 to CO2 and H2O. This allows 15 times more ATP to be made than that produced by glycolysis alone (Alberts, 2002). The discovery of the mitochondrion was gradual, however the term was derived from the greek words mitos (thread) and khondrion in 1898 by Carl Benda (Benda, 1898 544 /id); thus describing the structural double nature of this organelle. Decades ago, influential electron microscopy studies led to the dogmatic view of mitochondria as bean-shaped organelles. These studies revealed the ultrastructural hallmarks of mitochondria, which include double lipid membranes and unusual inner membrane folds termed cristae. Recent studies have led to renewed appreciation for the fact that the mitochondrial structure is highly dynamic. Imaging studies have revealed that mitochondria constantly move and undergo structural transitions; fusion and fission (Detmer and Chan, 2007b). 12
These processes are important as a number of diseases are directly caused by their dysfunction (Delettre, 2000; Alexander, 2000; Zuchner, 2004; Waterham, 2007). In this thesis mitochondrial form and function of healthy and diseased β-cells and intact islets as well as brown adipocytes is characterized. Furthermore, we test the hypothesis that mitochondrial form is of importance for function in these cell types. The results are presented in the context of the relevant literature. 2. Mitochondrial physiology 2.1. Mitochondria in a nut shell Mitochondria have multiple functions within the cell that include ATP-production, calcium signaling, and apoptosis. Each mitochondrion is bound by two lipid membranes. Together they create two separate mitochondrial compartments: the internal matrix space and the intermembrane space (Figure 1). The matrix contains hundreds of enzymes including those required for the citric acid cycle and β-oxidation. The inner membrane is folded into numerous cristae, which increase the surface area similar to intestinal villi. The cytochromes of the respiratory chain are arranged within the cristae. When cristae junctions open upon apoptotic stimuli cytochrome C may be released into the cytosol to cleave caspases (Alberts, 2002). The outer membrane is permeable to all molecules less than 5 kDa, however most can not pass through the inner membrane because of its high selectivity. The matrix also contains several copies of mitochondrial DNA (mtDNA), and various enzymes required for mitochondrial gene expression. Since each cell has hundreds of mitochondria, the mtDNA copy number is often above 1000/cell (Shay, 1990). The mitochondrial genome is essential for the respiratory function, however the majority of mitochondrial proteins are encoded by the nuclear genome and are then imported into the mitochondria (Wallace, 2005). The 16 kilobase circular mtDNA genome contains 37 genes. Thirteen of these genes encode protein subunits of respiratory complexes I, III, IV, and V; only complex II is composed of proteins encoded solely by nuclear genes (Wallace, 2005). The remaining genes encode transfer RNA (tRNA) and ribosomal RNA (rRNA) necessary for intramitochondrial protein synthesis (Anderson, 1981). Figure 1. Mitochondrial components: Image adapted from (Lamson and Plaza, 2002b). 13
2.2. Principles of bioenergetics Mitochondria are often referred to as “the powerhouse of the cell”; this because mitochondria produce approximately 90% of the ATP. ATP is produced by the respiratory chain complexes (I-V) (Figure 2). Complex I (ubiquinone NADH dehydrogenase) is responsible for the oxidation of NADH and pumps protons (H+) into the intermembrane space while reducing ubiquinone. Complex II (succinate dehydrogenase) oxidizes succinate into malate, thus liberating reducing equivalents (electrons) that are shuttled to complex III via ubiquinone. Complex III (ubiqinol- cytochrome-c reductase) receives electrons, liberating H+ in the process. Complex IV (cytochrome-c oxidase) reduces O2 to H2O, producing H+ in the process. As each complex moves electrons along the chain, protons are pumped out of the matrix into the intermembrane space. The proton gradient generated is used to drive ATP synthesis by Complex V (F1F0ATPsynthase), which phosphorylates ADP to ATP. This proton gradient is commonly referred to as the mitochondrial membrane potential (ΔΨm) and represents the energy available to drive changes in ATP/ADP ratio, and reactive oxygen species as well as controls mitochondrial calcium sequestration (Lowell and Shulman, 2005). The ΔΨm potential is considerably higher than the plasma membrane potential; ~150-180mV (Valdez, 2006) vs. ~60-90mV respectively (Wright, 2004). V Figure 2. The mitochondrial respiratory chain: The mitochondrial membrane potential (ΔΨm) is the electrochemical gradient that is present across the inner mitochondrial membrane and is generated by pumping H+ by complexes I,III,IV of the respiratory chain. This potential is used to drive ATP synthesis via complex V (F1F0ATPsynthase). Protons may also re-enter the matrix via other routes, e.g. via uncoupling proteins, without production of ATP. Complex I is considered as the major source of reactive oxygen species (ROS). Image adapted from (Lamson and Plaza, 2002a). During the transfer of electrons along the respiratory complexes, single electrons sometimes escape and result in a single electron reduction of molecular oxygen to form a superoxide anion (O2.-) (Fariss, 2005). It is estimated that as much as 1% of all oxygen consumed may result in the formation of reactive oxygen species (ROS) such as superoxide anions. The main sites of O2.- generation are at Complex I and the interface between ubiquinone and complex III (Nishikawa, 2000a). Although previously viewed as toxic byproducts it now appears that ROS may act as intracellular signaling molecules (Pi, 2007). 14
2.3. Mitochondrial oxygen consumption in cells and pancreatic islets 2.3.1. Assessing mitochondrial function A variety of cellular parameters may be used as indicators of mitochondrial function, including: the redox state of mitochondrial cytochromes; cellular ATP levels; ROS production; ΔΨm; and oxygen consumption. Although less widely measured, oxygen consumption is arguably the most informative of these parameters (Will, 2006), in that measurement allows a direct and specific assessment of the flow of the electron transport chain. The vast majority of cellular oxygen consumption is mitochondrial and occurs at complex IV in the respiratory chain as protons are pumped out into the intermembrane space (Mitchell, 1976). 2.3.2. Uncoupled and coupled respiration The majority of protons reenter the mitochondrial matrix through F1F0ATPsynthase and thereby fuel the conversion of ADP to ATP (Mitchell, 1976). Alternatively protons may reenter through mechanisms not coupled to ATP synthesis, “proton leak”, which stimulates “uncoupled respiration”. Therefore, oxygen consumption has a dual interpretation. Brand and colleagues have reported extensively on variations in proton leak among different cell types and even different species (Brand, 1991; Hulbert, 2002; Brand, 2003; Else, 2004; Jastroch, 2007; Lambert, 2007; Parker, 2008; Bottje, 2009). In principal, the level of uncoupled respiration is of interest as it reflects the cell’s bioenergetic efficiency. 2.3.3. Islet respiration In cell biology, oxygen consumption (or respiration) is measured either in isolated mitochondria, or in permeabilized or intact cells or tissue. Several assays for measuring oxygen consumption have been used over the years. In principal, assays have utilized either Clark type electrodes or fluorescent probes to report on media oxygen tension. The Clark electrode measures a flow of electrons, i.e. a current, that is dependent on the oxygen tension of the media (Clark, Jr., 1958). The fluorescent probes exhibit quenched fluorescence intensity emission in response to increased oxygen tension (Ji, 2002; Wu, 2007). Islets of Langerhans consist of several cell types including the β-cells (Figure 3). Mitochondria are essential for proper β-cells or islet function (Maechler, 2006). Mitochondrial metabolism of glucose derivatives is necessary for insulin secretion as described in Figure 3. Several different assays are used to measure islet respiration (Longo, 1991; Ortsater, 2000; Doliba, 2006; Papas, 2007c; Jung, 2008; Sweet, 2008c). Some have the advantage of having media flow-through and can sample in and outflow, thus e.g. enabling insulin secretion measurements. However, the flow-through methods can only run one sample per experiment and are cumbersome to use, i.e. user-dependent. Alternatively there are multiwell plates coated with oxygen sensitive fluorescent probes (Fraker, 2006; MacGregor, 2006). Though higher throughput, these plates do not provide detailed dynamic data, i.e. limited number of time points. Thus, up to now there has been 15
Figure 3. A) Islet of Langerhans from mouse. Islets comprise ~1% of the pancreas and each islet consists of ~1500 cells. Note the different cell types: β-cells (green; secrete insulin; 75% of cells), alpha cells (red; secrete glucagon; 19% of cells), delta cells (blue; 6% of cells). Adapted from (Brissova, 2005). B) Model for coupling of glucose metabolism to insulin secretion in the β-cell. Glucose is phosphorylated by glucokinase (GK) and converted to pyruvate (Pyr) by glycolysis. Pyruvate preferentially enters the mitochondria and fuels the citric acid cycle, resulting in the transfer of reducing equivalents to the respiratory chain, leading to hyperpolarization of the mitochondrial membrane (ΔΨm) and generation of ATP. Subsequently, closure of KATP-channels depolarizes the plasma membrane potential (ΔΨc). This opens voltage-gated Ca2+ channels, raising the cytosolic Ca2+ concentration ([Ca2+]c), which triggers insulin exocytosis. Several putative messengers, or additive signals, proposed to participate in the metabolism– secretion coupling are indicated. Adapted from (Maechler and Wollheim, 2001b) a lack of high throughput and user friendly assays that at the same time provide high quality data. In paper III the development of a novel approach to islet respirometry based on oxygen sensitive fluorescent probes and multiwell plates especially adapted for islets is presented. This assay can concurrently run 20 islet samples and test multiple conditions over a course of several hours. By applying drugs acting on the respiratory chain levels of basal, fuel-stimulated, uncoupled, maximal as well as non-mitochondrial respiration may be estimated under various conditions. 2.3.4. Significance of measuring islet respiration As mitochondrial dysfunction is much discussed in β-cell pathophysiology (Mulder and Ling, 2009), islet respiration methods are of importance for basic science. In addition, islet respirometry may also have a clinical use. Islet transplantation is a treatment under development for type 1 and advanced type 2 diabetes (Harlan, 2009). There are currently no reliable methods for assessing islet quality prior to transplantation (Papas, 2009). This is important, since many islet batches have quality issues, foremost because they come from critically ill donors and go through a traumatic treatment during their isolation. It was shown that islets with high oxygen consumption rates are more suitable for transplantation, at least to nude mice (Sweet, 2005; Papas, 2007a; Sweet, 2008b). The respirometry assay presented in paper III may therefore prove useful clinically to evaluate islet quality prior to transplantation. This because of its high troughput capability and its simple design that may make it easy to adapt by different transplantation centers across the world. 16
2.3.5. Uncoupled respiration of islets In clonal INS1 β-cells (derived from rat insulinoma) (Asfari, 1992), the proton leak or level of uncoupled respiration was reported to be four times higher than in clonal C2C12 myoblasts (Affourtit and Brand, 2008). However interesting, pancreatic islets function as a functional syncytium (Katzman, 2004) of ~1500 cells and results obtained on cell lines may not readily be extrapolated to the primary tissue. In paper III we present data showing that mouse islet mitochondria are highly uncoupled, ~60% of the basal respiration remains under oligomycin. This is the first characterization of uncoupled respiration in islet. Furthermore, paper III shows that INS1 cells exhibit levels of uncoupled respiration of ~40%. This data is in contrast with a previous INS1 cell study that measured uncoupled respiration to be 75% (Affourtit, 2008). The uncoupled respiration of C2C12 myoblasts (~20%) shown in paper III was similar to the previous study (Affourtit , 2008). Table 1 compares the rates of uncoupled respiration between different tissues and species. Interestingly the uncoupled respiration of islets appears higher than most other cells or tissues. System Percentage of Reference(s) ____________________________ total respiration ____________________ Rat β-cells (INS1) 40-75 (Affourtit, 2008), (Paper III) Mouse islets 60 (Paper III) Rat hepatocytes 20–26 (Rolfe, 1999),(Nobes, 1990)] Rat muscle 35–50 (Rolfe, 1996;Rolfe, 1999) Rat basal metabolic rate 20–25 (Rolfe, 1996;Rolfe, 1999) Mammal hepatocytes (mouse, ~20 (Porter and Brand, 1995) ferret, sheep, pig, horse) Avian hepatocytes (finch, Up to 21 (Else, 2004) sparrow, starling, currawong, pigeon, duck, goose, emu) Crocodile hepatocytes Up to 13–30 (Hulbert, 2002) Lizard hepatocytes Up to 30 (Brand, 1991) Frog hepatocytes Up to 20–25 (Brand, 2000) Lamprey hepatocytes Up to 25–50 (Savina, 1997) Snail hepatopancreas cells Up to 15–25 (Bishop and Brand, 2000) _______________________________________________________________ Table 1. Comparison of uncoupled respiration across different cell types and species. Table adapted and modified from (Brand, 1999) 2.3.6. Significance of uncoupled respiration The level of uncoupling is an important biological phenomenon as it reflects bioenergetic efficiency. Proton leak contributes to standard metabolic rate, i.e. energy consumption, by converting part of the mitochondrial proton gradient to heat. In fact, 16-31% of standard metabolic rate is caused by proton leak (Rolfe and Brand, 1996). The role of proton leak in human disease is still unknown, however some evidence suggest that it may increase with aging (Serviddio, 2007). Conversely caloric restriction, shown to increase lifespan, was shown to decrease proton leak (Bevilacqua, 2004; Johnson, 2006). Another study however found that long lived mice had higher levels of proton leak (Speakman , 2004). 17
Theoretically, if the fraction of coupled respiration could be increased, mitochondrial ATP-production may increase as a consequence. In β-cells this could in turn trigger higher insulin secretion as ATP stimulates closure of the ATP sensitive K-channels in the plasma membrane. Thus, the coupling efficiency of the respiratory chain of β-cells within pancreatic islets may represent a therapeutic target. On the other hand, it may be that the uncoupled respiration reflects essential processes for secretion. 2.3.7. Regulation of uncoupled respiration Proton leak, mirrored by uncoupled respiration, is typically divided into basal and inducible. Basal proton leak is present under resting conditions in all types of mitochondria that have been studied (Table 1) and may make a major contribution to metabolic rate. Inducible proton leak is, as the name implies, not present under resting conditions. A number of different mechanisms may contribute to proton leak. 2.3.7.1. Fuel regulation of uncoupled respiration In addition to glucose, free fatty acids and amino acids were shown to stimulate insulin secretion, either alone or as augmenters of GSIS (Newsholme, 2005). However, several additional messengers besides ATP are thought to play important roles (Maechler, 2006). In paper III we show that glucose, the free fatty acid palmitate and the amino acids leucine and glutamine in addition to increasing coupled respiration also dramatically increase uncoupled respiration. E.g., the combination of leucine and glutamine increased uncoupled respiration from ~60% to ~ 90%. The physiological role of fuel induced uncoupling is unclear; from a physiological perspective it appears inefficient. It may be that at the fuel levels tested in paper III mitochondrial ATP production is saturated although the metabolism preceding it is not. E.g., the citric acid cycle with its influxes from glycolysis, β-oxidation and amino acid metabolism may be working at a higher rate than the F1F0ATPsynthase. Uncoupling may act as “release valve”, diverting protons from F1F0ATPsynthase. The purpose of this higher rate may be to maintain a high production of amplifying signals from the preceding metabolism that augments insulin secretion. These may for example be generated from pyruvate shuttle traffic; NADPH, α-ketoglutarate and GTP (Jensen, 2008b) as well as citric acid cycle derived GTP (Kibbey, 2007). Further, it may be that uncoupling serves to protect the β-cell from fuel toxicity. Since the β-cell serve as a fuel sensor, it imports more fuel than is required for maintaining the ATP concentration. These fuels however may render them sensitive to fuel toxicity, as shown by studies on glucolipotoxicity (Poitout, 2008). Increased mitochondrial uncoupling may allow excess fuel to be turned into heat. 2.3.7.2. Lipid composition Some of the variation in proton leak between tissues and species may be explained by differences in mitochondrial inner membrane surface area. There is a correlation between mitochondrial proton conductance and the fatty acyl composition of inner-membrane phospholipids (Hafner, 1988; Brookes, 1997b; Hulbert, 2002; Brand, 2003). The content of n−3 polyunsaturates, particularly docosahexaenoate (C22:6,n−3), correlates with high proton conductance, and the content of monounsaturates, particularly oleate (C18:1,n−9), 18
correlates with low proton conductance. However, the proton conductance of phospholipid vesicles prepared from mitochondrial lipids is only 2–25% of the conductance of the mitochondria that they are derived from, and does not change when the composition changes (Brookes, 1997a; Brookes, 1997b). Consequently some other factor than membrane surface area or phospholipid composition must be an important determinant of the basal proton conductance of mitochondria. Naturally, as the mitochondrial membranes have numerous proteins these must be considered. 2.3.7.3. Uncoupling proteins To date two types of mitochondrial inner membrane proteins have been shown to be involved in proton leak; uncoupling proteins (UCP) and the adenine nucleotide translocase (ANT). UCP1 has long been recognized to mediate noradrenergic stimulated proton leak in brown adipocytes (Nicholls, 2001). The other uncoupling protein homologues, UCP2 and UCP3, are more controversial. UCP2 protein is mostly expressed in pancreatic islets, spleen, stomach, brain and lung while UCP3 is predominantly expressed in skeletal muscle, brown adipose tissue and heart (Chan and Harper, 2006). During the past years, UCP2 in islets have gained much attention. This interest was triggered by studies showing that UCP2 knock-out islets exhibit elevated ΔΨm and ATP- levels as well as increased insulin secretion (Zhang, 2001a). Furthermore, the UCP2 knock-out animals appeared protected against diet induced diabetes (Joseph, 2002). However, a recent study by Collins and colleagues has indicated that these findings may have been artifacts caused by the genetic background of the knock-out mice (Pi, 2009). UCP2 was knocked out in mice with three different strain backgrounds (C57BL/6J, A/J, 129/SvImJ). In contrast to previous studies, it was found that the insulin secretion was impaired (Pi, 2009). Furthermore, the knock-out islets showed high levels of oxidative stress including elevated levels of antioxidant enzymes and increased nitrotyrosine (Pi, 2009). Studies on β-cell lines have also showed opposing results. In INS1 β-cells with knock down of UCP2, it was calculated that 20% of the respiration was due to UCP2 (Affourtit, 2008). In addition insulin secretion was increased (Affourtit, 2008). On the other hand, another study where UCP2 was overexpressed found no alteration of uncoupled respiration (Galetti, 2009). Instead, decreased levels of oxidants was shown, thus in line with (Pi, 2009). Thus, the literature is somewhat contradictory. However there is little doubt that islet data is more relevant than clonal β-cell data. In paper III uncoupled respiration of mouse islets with a β-cells specific knock-out of UCP2 is examined. The generation of the knock-out animals was previously described elsewhere (Lee, 2009). We found no difference in uncoupled respiration as compared to control islets (paper III). In fact, UCP2 knock-out islets showed higher levels of basal respiration (paper III). These data add to the literature arguing that UCP2 is not primarily an uncoupling protein like its classic homologue UCP. Instead a primary role of UCP2 may be in regulation of ROS. 2.3.7.4. Adenine nucleotide translocase The other protein candidate for mediating protein leak, ANT, exchanges ADP for ATP across the mitochondrial inner membrane (Klingenberg, 2008). To examine the role of ANT in proton leak Brand and colleagues examined mice with knock-out of ANT1, an 19
isoform of ANT, and Drosophila melanogaster strains under- or overexpressing ANT (Brand, 2005). Skeletal muscle mitochondria were examined in the mice while whole body mitochondrial isolates were extracted from the flies. It was found that the amount of ANT present in the mitochondrial inner membrane strongly affects the basal proton leak. A major part of the leak appeared to be due to ANT (Brand, 2005). Similar results on the role of ANT were found in liver and brown adipocytes (Shabalina, 2006). It was suggested that ANT2 isoform may mediate fatty acid induced uncoupling while ANT1 may mediate a significant part of the basal proton leak. In paper III the role of the ANT is examined in mouse islets by using its specific inhibitor bongkrekic acid. The contribution of ANT to the basal level of uncoupled respiration was estimated to be ~31% (paper III). This is substantially lower than previously reported on isolated mouse skeletal muscle mitochondria where the ANT contribution was estimated to be between half to two-thirds of the basal proton conductance (Brand, 2005). Under fuel stimulated conditions the contribution of ANT to uncoupled respiration increased to 42%, likely due to increased nucleotide shuttling (paper III). The results on UCP2 and ANT are not surprising considering the different abundances of these proteins. ANT contributes 1–10% of total mitochondrial protein (Brand, 2005) while only 0.3%, 0.03% and 0.01% is contributed by the pyruvate carrier (Shearman and Halestrap, 1984;Paradies, 1984), UCP2 (Pecqueur, 2001) and UCP3 (Harper, 2002) respectively. In comparison to UCP2, UCP1 comprises 1–5% of mitochondrial protein in ◦ brown adipose tissue of mice kept below their thermoneutral temperature of 28 C (Stuart, 2001). Interestingly a recent study has shown that UCP2 content in pancreatic alpha cells is considerably higher than in β-cells (Diao, 2008b). Thus, it may be that UCP2 plays a greater role in alpha cells. 2.3.7.5. Permeability transition pore The permeability transition pore (PTP) is a large channel consisting of multiple subunits that increase mitochondrial inner membrane permeability to various solutes including protons (Rasola and Bernardi, 2007). The PTP classically opens in response to death stimuli and enables release of of cytochrome C that cleaves caspases (Rasola, 2007). However, PTP opening may also be partial and reversible (Liu and Murphy, 2009), and could thus in theory contribute to proton leak and a consequent increase in uncoupled respiration. Furthermore, one of the major components of the PTP complex is the ANT (Tsujimoto and Shimizu, 2007). PTP opening has been extensively examined in myocytes. In rat skeletal muscle exposed to anoxia/reoxygenation an increase in proton leak was demonstrated to be dependent on PTP (Navet, 2006). Interestingly, palmitate appeared to prevent this proton leak, probably because it caused uncoupling by itself. A study on mitochondria isolated from perfused rat hearts that were subjected to ischemia/reperfusion showed somewhat similar results (Nadtochiy, 2006). The increased proton leak after ischemia/reperfusion was inhibited to 50%, by carboxyattractyloside, an inhibitor of ANT, but also by cardioprotective treatments including the PTP inhibitor cyclosporin A. With these data in mind, the role of PTP in regulation of uncoupled respiration in islets was examined in paper IV by using cyclosporin A. No significant effect on the level of uncoupled respiration was found, both under low glucose and leucine/glutamine stimulation, thus pointing to no direct involvement of PTP. 20
2.3.7.6. Reactive oxygen species Traditionally, ROS have been thought of as useless by-products of respiratory metabolism in mitochondria and believed to be generally deleterious to biological systems (Finkel, 1998). However, ROS have emerged as physiological mediators of many cellular responses (Rhee, 2006) and some evidence suggests that these molecules may serve a signaling function (Pi, 2007). In β-cells, it was recently suggested that low levels hydrogen peroxide derived from glucose metabolism serves as a signal for insulin secretion, whereas oxidative stress may disturb its signaling function (Pi, 2007; Pi, 2009). Nearly all of ROS are produced in mitochondria because of interaction of oxygen with free electrons released by the respiratory chain. Figure 4. Principles of reactive oxygen species (ROS) generation. ROS are foremost generated by the mitochondrial respiratory chain where the majority of oxygen is consumed. SOD and catalase are important scavengers of superoxide. (O2· ¯, superoxide anion radical; H2O2, hydrogen peroxide; ·OH, hydroxyl radical; ONOO-, peroxynitrite; SOD, superoxide dismutase, NO, nitric oxide). Adapted from (Kyaw, 2004) ROS was shown to induce uncoupling (Echtay, 2002). One mechanism by which ROS does so is by lipid peroxidation that leads to production of reactive aldehydes such as 4- hydroxynonenal (Echtay, 2003). These aldehydic lipid peroxidation products are able to modify proteins such as mitochondrial uncoupling proteins and the ANT, converting them into active proton transporters. Furthermore, in β-cells ROS was described to increase with fuel exposure (Pi., 2007) (paper III). In addition fuels such as palmitate may induce uncoupling in islets (Carlsson, 1999). Considering these data, the effect of scavenging ROS with the superoxide-dismutase mimetic tetrakis (4-benzoic acid) porphyrin manganese(III) (TBAP) on islet respiration was examined in paper III. Interestingly, TBAP totally abolished the fuel stimulated increase in uncoupled respiration. This suggests that ROS may be the molecular link between fuel metabolism and uncoupled respiration in islets. Furthermore, in paper III it is shown that the ANT uncoupling activity increases under fuel stimulated conditions. It was previously shown that in aging ANT is specifically targeted by ROS (Yan, 1998), suggesting that ANT may have a particular sensitivity to ROS. A possible mechanism for how ROS induce uncoupling in the islets may be by stimulating ANT. 2.4. Mitochondrial dynamics The mitochondrial morphology is a dynamic property that can form a variety of shapes: from long, interconnected tubules to individual small spheres (Frazier, 2006). In 21
eukaryotes, the overall morphology of the mitochondria is maintained by balancing the opposing processes of mitochondrial fusion and fission, collectively termed mitochondrial dynamics. These processes not only control mitochondrial morphology but also play an important role in mitochondrial function (Detmer and Chan, 2007a; Detmer, 2007b). Without mitochondrial dynamics, the mitochondrial population consists of autonomous organelles that have impaired function that include reduced metabolism, and increased apoptosis (Chan, 2006). 2.4.1. Regulation of mitochondrial fusion Mitochondrial fusion is defined as the merger of two mitochondria resulting in one larger mitochondrion (Figure 5) (Skulachev, 2001; Detmer, 2007a; Detmer, 2007b). In mammalian cells fusion is regulated by three known transmembrane GTPases: mitofusin- 1 (Mfn1), mitofusin-2 (Mfn2) and Optic Atrophy 1 (Opa1). Opa1 is located within the inner mitochondrial membrane (Olichon, 2003). Mfn1 and 2 are located within the outer mitochondrial membrane. Mfn1 and Mfn2 appear to play similar roles in mitochondrial fusion although Mfn1 requires Opa1 for its function while Mfn2 does not (Zhang and Chan, 2007). Deficiency of either protein results in mitochondrial fragmentation. Mfn1 and 2 can functionally replace each other. Cells lacking Mfn1 can be rescued by overexpression of Mfn2; conversely cells lacking Mfn2 can be rescued by overexpression of Mfn1 (Chen, 2003). Moreover, Mfn-null cells can be fully rescued by overexpression of either mitofusin (Chen, 2003). 2.4.2. Regulation of mitochondrial fission Mitochondrial fission is the division of a mitochondrion to form two or more separate mitochondrial units (Figure 6) (Yoon, 2004). In eukaryotes mitochondrial fission involves the transmembrane protein Fis1 and GTPase dynamin-related protein (Drp1). Drp1 is a key component of the mitochondrial fission machinery. A minor fraction of Drp1 is localized to punctate spots on mitochondrial tubules, and a subset of these spots Figure 5. Fusion event: Mitochondrial fusion consists of outer membrane (OM) fusion followed by inner membrane (IM) fusion (top panel). Mfn1 and Mfn2 localized on the outer mitochondrial membrane tether with the inner mitochondrial protein, Opa1 resulting in mitochondrial fusion (bottom panel). Adapted from (Detmer, 2007). 22
mark future sites of fission. The majority of Drp1 is free in the cytosol and is recruited by Fis1 upon induction of fission (Smirnova , 2001). Drp1 activity is inhibited by phosphorylation by protein kinase A which in turn is regulated by cAMP (Cribbs, 2007).Inhibition of Drp1 by expression of a dominant-negative (DN) mutant leads to increased length and interconnectivity of mitochondrial tubules (Lee , 2004). ` Figure 6. Fission event: Fis1 is localized on the outer mitochondrial (OM) membrane, Drp-1 binds at scission sites upon which fission occurs resulting in two individual mitochondria. Adapted from (Chen, 2005). 3. Mitochondrial heterogeneity Over the past years, it has been shown in numerous studies that mitochondria display functional and structural heterogeneity. Any measured biological parameter varies to some degree. When the magnitude of the variations respond to physiologically and pathologically relevant alterations it may by itself become the parameter of interest. Mitochondrial subcellular heterogeneity is altered by metabolic stress (paper I) and apoptosis (D'Herde , 2000;Krysko , 2001), and therefore deserves attention. Recent years’ progress in the understanding of the mitochondrial life cycle, including mitochondrial dynamics and mitochondrial autophagy (mitophagy) suggest that the two processes control and maintain the extent of heterogeneity. Conditions that increase heterogeneity also affect mitochondrial dynamics and autophagy. In cell death, both heterogeneity (Salvioli , 2000;Krysko , 2001) and autophagy (Kroemer and Levine, 2008) increase and mitochondrial dynamics is inhibited (Suen , 2008). In pancreatic β-cells, metabolic stress in the form of high levels of glucose and free fatty acids disrupts mitochondrial dynamics (paper II), increases heterogeneity (paper I) and upregulates autophagy (Choi , 2008). 3.1. Subcellular heterogeneity Mitochondrial heterogeneity has been reported from a diverse range of primary cells and cell lines including neurons, myocytes, exocrine and endocrine cells, as well as from brown adipocytes (paper IV). A variety of techniques have been used; however, imaging data is dominating the literature. A consensus definition of mitochondrial heterogeneity is lacking in the literature. In principle, heterogeneity is the variance in the measured parameter representing the combined effects of the diversity of the sampled individuals and the noise introduced by the sampling technique. When considering the functional significance of heterogeneity one needs to determine what portion of it is contributed by the noise generated by the methodology of data acquisition and analysis. For example, in the case of confocal microscopy large variance can be generated by the fact that mitochondria are not all in the same focal plane; therefore, the intensity of the 23
fluorescence signal largely depends on the position of each mitochondrion in the z-axis. In addition certain image processing algorithms (such as filters) may widen or narrow the spectrum of the signal. The functional significance of heterogeneity can be determined by demonstrating that heterogeneity is altered by relevant effectors. For example, the span of ΔΨm heterogeneity is altered in pancreatic β-cells responding to glucose stimulation (paper I). In addition, the temporal characteristic of the heterogeneity is also of importance. 3.1.1. Mitochondrial membrane potential heterogeneity ΔΨm is a widely used bioenergetic parameter affecting multiple mitochondrial functions including ATP synthesis, Ca++ sequestration, protein import, mitochondrial fusion, mitochondrial autophagy, and the generation of reactive oxygen species (ROS) (Nicholls , 2000). ΔΨm is regulated by factors contributing to its build up, e.g. fuel input and respiratory chain activity and factors contributing to its dissipation, e.g. F1F0-ATPase activity as well as uncoupling mechanisms and other ion fluxes than protons (Huser, 2000). Within a physiologically relevant ΔΨm range, the maximum ATP/ADP ratio that can be maintained by mitochondria decreases by up to 10-fold for every 14mV decrease in ΔΨm, thus the level of ΔΨm reflects a mitochondrion’s energetic capacity (Nicholls, 2004). Glucose-induced ΔΨm hyperpolarization correlates well with induction of insulin secretion (dependent on raises in ATP/ADP ratio) (Heart , 2007) as well as with increased mitochondrial oxygen consumption as shown in paper III. The oxygen consumption data indicates that glucose-induced ΔΨm hyperpolarization is due to increased proton pumping activity by the electron transport chain. However, in cases where oxygen consumption data is not available, caution should be practiced when interpreting changes in ΔΨm. For example, while depolarization is frequently attributed to mitochondrial respiratory dysfunction, it should be kept in mind that increased ATP- production may lead to depolarization under some circumstances, e.g. in state 3 respiration of isolated mitochondria where ADP, the substrate for oxidative phosphorylation and a possible dissipater of ΔΨm, is in excess. As in situ mitochondria cannot be accessed directly, indirect methods which predominantly use membrane- permeate cationic fluorescent dyes have been employed to monitor ΔΨm in cells by imaging. While these dyes have been widely used they present a number of challenges including unspecific binding, photo-toxicity and interference with cell metabolism. To appropriately interpret data generated using ΔΨm probes these confounding factors should be controlled for. Unspecific binding can be estimated by adding a mitochondrial uncoupler, which would dissipate the ΔΨm and lead to loss of mitochondrial staining. Photo-toxicity can be reduced by lowering the intensity of the excitation light during imaging. No matter what intensity is finally being used, the researcher has to address the possibility of photo-toxicity by monitoring the effect of dye excitation on a relevant biological function that can be accepted as a measure of cellular viability and function. By using low dye concentrations metabolic interference can be minimized and controlled for by testing cellular function in the presence of the dye. When imaging mitochondria, it is also important to be aware of the cell’s z-axis. A difference in dye fluorescence intensity between two different mitochondria in an individual image may arise either from a real difference in dye concentration (and thus ΔΨm) or alternatively from the mitochondria being in different focal planes. This may be corrected for either by using 24
confocal microscopy z-stack imaging, where multiple focal planes are recorded and compiled into a single image, or by combining two fluorescent probes and using their ratio for assessing ΔΨm as described in paper I. Moreover, ratio imaging can be used to account for artifacts stemming from the limitations of image resolution (200 nm in conventional confocal microscopes). For example, a mitochondrion may cover only part of the factual area that an image pixel “records” from. The light intensity emitted from the mitochondrion will therefore be divided by a larger area than it is actually covering, leading to a false impression of reduced dye concentration. In summary, data generated using ΔΨm probes should be interpreted with caution and the availability of appropriate controls in which potential artifacts are accounted for should be verified before conclusions are made. Here follows a review of the results and caveats of the main probes used to study heterogeneity in ΔΨm. 3.1.2. JC-1 Tetrachloro-1,1,3,3-tetraethylbenzimidazol-carbocyanineiodide (JC-1) is a fluorescent dye that at low ΔΨm exists as green-emitting monomers but when ΔΨm increases, forms red emitting aggregates (Smiley , 1991). The proportion of aggregates to monomers in the inner membrane of the mitochondria is dependent both on the membrane potential as well as on the concentration of the dye in the cytosol. JC-1 has been used to show ΔΨm heterogeneity in a variety of cells, including intact human fibroblasts (Smiley , 1991); HeLa cells and hepatocytes (Collins , 2002); mouse oocytes and blastocysts (Van , 2003; Van , 2006; Van and Davis, 2006); mouse and human early embryos (Acton , 2004); pancreatic β-cells (paper I); human astrocytes, HEp-2, MDCK and Vero cells (Diaz, 1999); rat cardiomyocytes (Bowser , 1998); as well as in isolated liver mitochondria (Cossarizza , 1996). Although important in revealing heterogeneity, JC-1 possesses a number of drawbacks, of which the most important is that its partition to the inner membrane is not Nernstian and is therefore not reliable for calculating ΔΨm according to the Nernst equation (Nicholls , 2000). Consequently, it has only been used in qualitative descriptions of ΔΨm heterogeneity. Moreover, if staining exceeds 30 min, or slightly higher concentrations are used, JC-1 may produce peculiar artifacts (Wikstrom, 2009). In some cells, JC-1 may aggregate into long nail-like structures, very different from mitochondrial architecture. JC-1 appears to load more readily into projections of cells such as neurons and INS1 β-cells, probably reflecting faster dye loading due to the projections’ higher plasma membrane to cytosol ratio. Thus, JC-1 also has drawbacks of qualitative nature. 3.1.3. TMRE/TMRM allows for quantitative evaluation of heterogeneity Tetramethylrhodamine-ethyl-ester (TMRE) and methyl-ester (TMRM) are two similar ΔΨm probes. They have the advantages of comparatively low mitochondrial toxicity. Their partition to the mitochondria inner membrane is Nernstian, enabling quantitative studies of ΔΨm heterogeneity (Nicholls , 2000). A number of studies have used these dyes. (Distelmaier , 2008) reported on ΔΨm heterogeneity in primary human skin fibroblasts. In primary cultures of human fetal astrocytes and adult fibroblasts, longitudinal profiles of single mitochondria were homogenous, while the ΔΨm between mitochondria differed (Diaz , 2000). Distribution curves of ΔΨm of mitochondrial populations were shown in several reports (Loew , 1993; Zhang , 2001b) and is also 25
demonstrated in paper I). The similarity in ΔΨm heterogeneity measured from neuroblastoma and primary pancreatic β-cells is remarkable, the mean standard deviation in ΔΨm, as expressed in mV, is 11mV in neuroblastoma cells (Loew , 1993) and 9mV in primary pancreatic mouse β-cells (paper I). These subcellular variances in ΔΨm may appear small but may translate to large differences in ATP production between mitochondria within the same cells (Nicholls, 2004). 3.2. Heterogeneity in compromised cells While mitochondrial heterogeneity appears to be a universal phenomenon, increased levels of heterogeneity have been associated with cell pathology. In mouse pancreatic β- cells ΔΨm heterogeneity is increased when cells are metabolically stressed with high levels of glucose and free fatty acids (paper I). The effect of stress on mitochondrial heterogeneity was tested in an ischemia-reperfusion model (Kuznetsov, 2004a; Kuznetsov , 2004b; Kuznetsov , 2004c; Kuznetsov , 2006) It was shown that mitochondria in rat cardiomyocytes are heterogeneous in terms of ΔΨm, Ca++, ROS and flavoproteins and that heterogeneity increased after cold ischemia-reperfusion. This effect may be attributed to heterogeneity in PTP opening and/or cytochrome C release (Kuznetsov , 2004b), however this was not directly tested. PTP opening was attributed to ROS, since it was prevented by antioxidants. In mitochondria isolated from ischemic rabbit hearts, it was shown that subsarcolemmal (SS), but not intermyofibrillar (IMF) mitochondria, had a large decrease in oxidative phosphorylation, likely due to decrease in cytochrome C content (Lesnefsky , 1997). It was further shown in quail apoptotic granulosa cells, that cytochrome C release as well as ΔΨm are heterogeneous, and it was suggested that ATP needed for completion of the apoptotic cascade may be generated in a subset of still respiring mitochondria (D'Herde , 2000; Krysko , 2001). Moreover, by tagging cytochrome C with GFP Heiskanen et al. demonstrated that those mitochondria that were found to be depolarized in staurosporine-treated apoptotic rat pheochromocytoma cells indeed lose their cytochrome C, thus further supporting PTP opening and cytochrome C release as causing ΔΨm heterogeneity in apoptosis. In HeLa cells stained with calcein-AM and exposed to oxidative stress in the form of peroxide, it was shown that the loss of calcein-AM from mitochondria was heterogeneous, which is likely to reflect heterogeneous PTP opening (Collins , 2002). Furthermore, in unstressed cells PTP may operate in a reversible low conductance mode and not be associated with cell death (Ichas , 1997). This low conductance mode may be implicated in depolarization prior to autophagy (Kim , 2007). Thus it is possible that heterogeneity in healthy cells may also be affected by heterogeneous PTP activity. Finally, in cells that are rapidly dividing, a fraction of the cells examined may be undergoing mitosis. In early mitotic phase in HeLa cells, mitochondria undergo transient fragmentation (Taguchi , 2007). Thus, the proliferation rate may influence mitochondrial heterogeneity measurements to some degree. 3.3. Mechanism of mitochondrial heterogeneity Heterogeneity in mitochondrial function may be attributed to intrinsic and extrinsic sources. In principle, the metabolic functions of a mitochondrion may be influenced by external signals from the cytosol or other organelles and thus be dependent on its location in the cell. Alternatively, function may depend on intrinsic properties or content of the 26
You can also read